Emotions influence our lives. Observational methods of measuring affective behavior have yielded critical insights, but a persistent barrier to their wide application is that they are labor-intensive to learn and to use. An automated system that can quantify and synthesize human affective behavior in real-world environments would be a transformational tool for research and for our everyday life. Are we there yet?

Recent breakthroughs in automated multimodal analysis and synthesis make possible objective, repeatable, efficient measure of emotion-relevant behavior in naturalistic environments. In this talk I will present my work on: (1) novel deep learning based computational approaches to measure and synthesize affective behavior; (2) their applications to realize closed-loop adaptive Deep Brain Stimulation (aDBS) systems for non-motor neuropsychiatric disorders and to gain new understanding of the neural and social bases of emotion; and (3) current challenges that affective computing faces in real-world conditions.

Bio: László A. Jeni is Project Scientist in the Robotics Institute at Carnegie Mellon University, Pittsburgh, PA, USA. He specializes in computational behavior science, specifically in areas of modelling, understanding, and synthesizing human behavior using diverse sensors. He received his Ph.D. in 2012 from the University of Tokyo, Japan. He worked as a Senior Computer Vision Specialist at Realeyes – Emotional Intelligence, before joining the Robotics Institute. To his credit he has over 40 peer reviewed publications. His honors include Best Paper Awards at the IEEE Conference on Human System Interaction (HSI’2011) for work on validating observations of human activity and at the Conference on Automatic Face and Gesture Recognition (FG’2015) for work on dense 3D face alignment from 2D video. He has organized workshops and challenges on 3D Face Alignment in the Wild at ECCV and on Facial Expression Recognition and Analysis at FG.